Odgovor humanističkih nauka na nove tehnologije: da li će mašinsko učenje promeniti časove jezika i književnosti?
Response of humanities to new technologies: will machine learning affect language and literature classes?

Author(s): Teodora Todorić Milićević
Subject(s): Anthropology, Social Sciences, Education
Published by: Етнографски институт САНУ
Keywords: humanities; higher education; literature; chatGPT; machine learning
Summary/Abstract: Since the last couple of decades of the 20th century higher education in most countries has undergone extensive changes. These are, for instance, an increase in the number of students; development of information technologies; globalization and dominance of the English language in science; the rise and spread of neoliberal ideology and hence the emergence of a market oriented university. These changes have caused (self)re-examination of the humanities, which are often criticized for the lack of a practical purpose and for providing irrelevant and non-instrumental knowledge. The aforementioned changes have impact on the Republic of Serbia as well and are particularly noticeable in the problems faced by the Faculty of Philology at the University of Belgrade, which faces decline in student interest in Serbian language and literature majors. At the end of November 2022, higher education faced a new challenge – an advanced chatbot from the company OpenAI called ChatGPT was released. This free software quickly generates answers to various user questions and writes essays, emails and various types of texts. Its emergence has sparked discussions about ethics, plagiarism in science, methodology and, finally, the purpose of certain disciplines. This paper, using the example of literature, analyses answers which that discipline can give in response to new technologies. We point out that literature – as it should be taught in colleges and then taught in middle and elementary schools – has different purpose than machine learning-based tools. In fact, the competencies that pupils and students should acquire in literature classes represent insurmountable obstacles for machine learning, which are creative reasoning in new contexts, the art of interpretation and the ability to make moral judgments.

• Details
• Contents